Accepted Manuscript Characterization and simulation on antireflective coating of amorphous silicon oxide thin films with gradient refractive index Lu Huang, Qi Jin, Xingling Qu, Jing Jin, Chaochao Jiang, Weiguang Yang, Linjun Wang, Weimin Shi PII:
S0749-6036(16)30239-7
DOI:
10.1016/j.spmi.2016.05.026
Reference:
YSPMI 4348
To appear in:
Superlattices and Microstructures
Received Date: 11 April 2016 Revised Date:
16 May 2016
Accepted Date: 19 May 2016
Please cite this article as: L. Huang, Q. Jin, X. Qu, J. Jin, C. Jiang, W. Yang, L. Wang, W. Shi, Characterization and simulation on antireflective coating of amorphous silicon oxide thin films with gradient refractive index, Superlattices and Microstructures (2016), doi: 10.1016/j.spmi.2016.05.026. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Characterization and simulation on antireflective coating of
2
amorphous silicon oxide thin films with gradient refractive index
3
Lu Huang
4
Weiguang Yang a, Linjun Wang a, Weimin Shi a
, Qi Jin a, Xingling Qu a, Jing Jin a, Chaochao Jiang a,
RI PT
a,*
5 6
a
7
China
SC
School of Materials Science and Engineering, Shanghai University, Shanghai 200444,
9 10
*corresponding author: Tel. numbers: +86 21 66138066, E-mail address:
[email protected] (L. Huang)
11
ABSTRACT
TE D
12
M AN U
8
The optical reflective properties of silicon oxide (SixOy) thin films with gradient
14
refractive index are studied both theoretically and experimentally. The thin films are
15
widely used in photovoltaic as antireflective coatings (ARCs). An effective finite
16
difference time domain (FDTD) model is built to find the optimized reflection spectra
17
corresponding to structure of SixOy ARCs with gradient refractive index. Based on the
18
simulation analysis, it shows the variation of reflection spectra with gradient
19
refractive index distribution. The gradient refractive index of SixOy ARCs can be
20
obtained in adjustment of SiH4 to N2O ratio by plasma-enhanced chemical vapor
21
deposition (PECVD) system. The optimized reflection spectra measured by
22
UV-visible spectroscopy confirms to agree well with that simulated by FDTD method.
AC C
EP
13
ACCEPTED MANUSCRIPT 23
Keywords: antireflective coating; amorphous silicon oxide; gradient refractive index;
24
finite difference time domain
25
1. INTRODUCTION
RI PT
26
In the recent years, amorphous silicon (α-Si) thin films on glass substrates have
28
attracted a great deal of attention in photovoltaics, microelectronics and display
29
technologies because of its potential applications for electronic devices, especially
30
thin film solar cells [1-3]. Efficient solar cells must satisfy two kinds of requirements:
31
both strongly optical absorption and effectively electrical transportation. To enhance
32
performance of antireflection and passivation, dielectric oxide layers must be selected
33
to deposit on the surface of α-Si thin films. The antireflective coatings (ARCs) have
34
been widely used in manufacturing process of conventional crystalline solar cells
35
(c-Si) to benefit optical absorption, such as silicon nitride (SixNy) deposited on the
36
surface of n-type c-Si [4-6] and aluminum oxide (AlxOy) prepared on the surface of
37
p-type c-Si [7-9]. Because of the match of lattice to α-Si substrates, amorphous silicon
38
oxide (SixOy) thin films have been used to prepare ARCs by plasma-enhanced
39
chemical vapor deposition (PECVD) [10-11], magnetron sputtering [12-13] and
40
sol-gel methods [14-16]. Compared to physical vapor deposition (PVD) method, it is
41
more suitable to prepare multilayer films with different refractive index by PECVD
42
method. The average reflectance for double-layer ARCs are lower over a broader
43
wavelength range than for a single-layer ARC, because single-layer ARC has only
44
minimal point of reflectance [17-18]. With these requirements, double-layer SixOy
AC C
EP
TE D
M AN U
SC
27
ACCEPTED MANUSCRIPT ARCs with gradient refractive index not only decrease electrical recombination
46
effectively, but also enhance optical antireflection strongly [19-20]. However, it is
47
worth discussing how to enhance antireflective effect of α-Si thin film solar cells
48
through different distribution structures of SixOy ARCs.
RI PT
45
In this paper, we present theoretical and experimental study of optical reflective
50
properties of SixOy ARCs with gradient refractive index. The finite difference time
51
domain (FDTD) method is used to study the light-modulated characteristics of ARCs
52
with gradient refractive index. Reflection spectra for different distribution structures
53
of SixOy ARCs with gradient refractive index are simulated by FDTD method. This
54
experimental work aims to focus on the correlation between reflection spectra and
55
gradient refractive index distribution. Surface characterization and FDTD simulation
56
analysis of SixOy ARCs with gradient refractive index are investigated simultaneously.
58
M AN U
TE D
57
SC
49
2. EXPERIMENTAL DETAILS
The hydrogenated amorphous silicon oxide (SixOy:H) thin films are prepared by
60
the following procedure. First, SixOy:H films as a layer is deposited on the glass
61
substrates using PECVD by silane (SiH4), nitrous oxide (N2O) and hydrogen (H2).
62
The experimental parameters of deposited SixOy:H thin films are shown in Tab.1. The
63
PECVD system is HLF-400 made in Beijing Beiyi Innovation Vacuum Technology
64
Co. Ltd. Before the thin films are deposited on glass substrate, heat-resistant tape has
65
been adhered to the corner of glass substrate. After we finish depositing SixOy:H thin
66
films and uncover tape from the substrate, the thickness of thin film is measured on
AC C
EP
59
ACCEPTED MANUSCRIPT the step between glass substrate and thin films by stylus profiler. Thickness
68
measurement of stylus profiler is Vecco Dektak 150. Then double-layer structures are
69
formed on the surface of first layer according to above procedure again. All of
70
samples have been dehydrogenated, after thin films were annealed in vacuum
71
condition at 200°C for 30 minutes.
RI PT
67
The structural orientation of SixOy films are measured by a D/Max-2200 X-ray
73
diffractometer (XRD) of Rigaku using the Cu Kα radiation. The cross-section
74
morphology and element composition of thin films is characterized respectively by
75
scanning electron microscope (SEM) and energy dispersive spectrometer (EDS) made
76
by JEOL JSM-6700F. The refractive index and thickness are measured by
77
spectroscopic ellipsometer of Sentech SE-400adv. UV-visible reflection spectra are
78
performed by Hitachi U-2910 spectrophotometer.
M AN U
TE D
79
SC
72
3. RESULTS AND DISCUSSION
81
3.1 Structural modeling
EP
80
In this paper, the software of FDTD Solution is used as analysis for simulating
83
different distribution structures of SixOy ARCs with gradient refractive index. The size
84
of rectangular model is that the length, width and height are 2.5 µm, 2.5 µm and
85
0.1µm respectively. The physical and optical values of silicon oxide materials are
86
used as parameters in the simulation. Different structures of SixOy ARCs with gradient
87
refractive index are selected to simulate the properties of optical reflection. The
88
optimal results can be analyzed to obtain the lowest average reflectance in the
AC C
82
ACCEPTED MANUSCRIPT 89
wavelength range of visible light. The basic cross-section structures of single ARC and double ARCs on the
91
substrate have been shown in Fig. 1(A) and Fig. 1(B) respectively. The reflectance of
92
incident light of wavelength λ from the surface of substrate covered by a single
93
non-absorbing layer is given by Eq. 1:
r12 + r22 + 2r1r2 cos 2θ R= 1 + r12 r22 + 2r1r2 cos 2θ
where r1 and r2 are the individual Fresnel reflectance given by Eq. 2:
r1 =
96
97
n0 − n1 n −n , r2 = 1 s n0 + n1 n1 + ns
M AN U
95
[1]
SC
94
RI PT
90
[2]
and θ is the phase difference of the single layer, β is refractive angle given by Eq.3:
2θ =
4πn1d1 λ cos β
[3]
TE D
98
And n1 is the refractive index of the thin film, d1 is its thickness, and ns is the
100
refractive index of the substrate. The reflectance exists minimum at a quarter
101
wavelength where n1d1 = λ0 / 4 and for odd multiples of λ0 / 4 . This minimal value
102
is given by Eq. 4:
104
AC C
103
EP
99
R = [( n1 − n0 ns ) ( n1 + n0 ns )]2 2
2
[4]
And R equals zero if it meets Eq. 5. Since n0 equals one for air medium, the
105
refractive index of the film should equal the square root of index of the substrate for
106
zero reflectance. The reflective value becomes higher, if it is higher or lower than this
107
quarter wavelength value due to variations in the refractive index with wavelength.
108
n1 = ( n0 × ns )1 2
109
Lower average reflectance can be obtained using double ARCs instead of single
[5]
ACCEPTED MANUSCRIPT 110
layer, where the inner layer next to the substrate and the outer layer upon the inner
111
layer. The double layer system is a better match between the high refractive index of
112
the substrate and the low refractive index of air. The reflectance of light from the
113
surface of the structure is given by Eq. 6: r12 + r22 + r32 + r12 r22 r32 + 2r1r2 (1 + r32 ) cos 2θ1 + 2r2 r3 (1 + r12 ) cos 2θ 2 R=
+ 2r1r3 cos 2(θ1 + θ 2 ) + 2r1r22 r3 cos 2(θ1 − θ 2 ) 1 + r12 r22 + r12 r32 + r22 r32 + 2r1r2 (1 + r32 ) cos 2θ1 + 2r2 r3 (1 + r12 ) cos 2θ 2 + 2r1r3 cos 2(θ1 + θ 2 ) + 2r1r22 r3 cos 2(θ1 − θ 2 )
[6]
r1 =
116
SC
where r1, r2 and r3 are the individual Fresnel reflectance given by Eq. 7:
n0 − n1 n − ns n − n2 , r2 = 1 , r3 = 2 n0 + n1 n1 + n2 n2 + n s
M AN U
115
RI PT
114
[7]
117
where ns is now the index of the substrate, θ1, θ2 are the phase difference and β, γ are
118
refractive angle of the double layers given by Eq. 8:
2θ1 =
121
122
[8]
The reflectance has either a minimum or a local maximum for quarter wavelength optical films ( n1d1 = n2 d 2 = λ0 / 4 ). This reflectance is given by Eq. 9:
R = [(n12 ns − n0 n22 ) (n12 ns + n0 n22 )]2
EP
120
4πn1d1 4πn2 d 2 ,2θ 2 = λ cos β λ cos γ
TE D
119
[9]
2 2 Which R approaches zero if the condition meets n2 / n1 = ns n0 , and
124
approaches a local maximum with zero reflectance on either side if the condition
125
meets n1n2 = n0 ns . Therefore the average reflectance for double ARCs are lower
126
over a broader wavelength range than for a single ARC, because single ARC has only
127
minimal value of reflectance [21].
AC C
123
128 129 130
3.2 Morphology and structure We select ten kinds of samples deposited by different gas ratio of SiH4 to N2O,
ACCEPTED MANUSCRIPT 131
such as 1:0.5, 1:1, 1:2 and 1:3 respectively, to investigate the cross-section
132
morphology of thin films. The related experimental and measured parameters of
133
samples for single- and double-layer films have been shown in Tab. 2. The characterization of XRD patterns is also displayed in Fig.2. In the picture,
135
SixOy and Si phase can be obtained through different peak shift. The SixOy exhibits a
136
broad peak at ~28° (sample c in Fig.2). Because SixOy thin films still includes parts of
137
amorphous phase, the broad XRD peak at angle ~28° demonstrates amorphous phase.
138
When the ratio of SiH4 in mixed gas is increased, a new peak of crystalline silicon
139
(c-Si) appears at angle ~30° corresponding to the standard XRD cards for crystalline
140
Si (111) (sample a and sample b in Fig.2).
M AN U
SC
RI PT
134
The cross-section SEM image of sample h has been demonstrated in Fig. 3(A),
142
indicating that the SixOy film thickness of sample h is about 112 nm. The double SixOy
143
ARCs have been deposited on the glass substrate coating Sn2O3:In transparent
144
conductive oxide (ITO), because SixOy layer and glass substrate can be separated
145
obviously by ITO layer from SEM image. The atomic ratio of element Si to O (x/y) is
146
about 56.59/43.41=1.3, which is characterized by EDS analysis in Fig. 3(B). As a
147
result, sample h is proven to be the best condition of SixOy ARC with gradient
148
refractive index, which can also be confirmed by optical results experimentally and
149
numerically later.
AC C
EP
TE D
141
150 151 152
3.3 Optical properties The refractive index and thickness of SixOy thin films measured by spectroscopic
ACCEPTED MANUSCRIPT ellipsometer have been listed in Tab. 2. Especially the correlation between refractive
154
index and gas ratio of single SixOy thin films has been shown in Fig. 4. When the ratio
155
of SiH4 in mixed gas is increased, the velocity of growth for SixOy thin film becomes
156
faster and the value of refractive index for thin film becomes higher, because the
157
increment of SiH4 ratio is benefit to enhance the thickness of SixOy thin film in the
158
same deposition time. It is found that the gradient refractive index of SixOy ARCs can
159
be obtained by adjustment of SiH4 ratio in mixed gas.
SC
RI PT
153
In the structure of ARC on the substrate, the reflection spectra are analyzed by
161
FDTD simulation. The reflection spectra are assumed to vary through different
162
refractive index of single and double layer thin film. The parameters of refractive
163
index and thickness in Tab. 2 are originated from spectroscopic ellipsometer. It is
164
proved theoretically that the correlation between reflection spectra and gradient
165
refractive index distribution can be obtained from FDTD simulation. Meanwhile the
166
reflection spectra of SixOy ARCs with gradient refractive index are deduced from the
167
UV-visible spectrophotometer. We obtained reflection spectra by UV-visible spectral
168
characterization as illustrated in Fig. 5(A) and FDTD simulation as shown in Fig.
169
5(B). The reflection spectra corresponding to sample a, h, i and j have been displayed
170
in Fig. 5 respectively.
TE D
EP
AC C
171
M AN U
160
When the substrate is silicon material, the optimized value of refractive index for
n0 ns = 3.9 ≈ 2 by Eq.5. To obtain lowest average
172
single ARC is about
173
reflectance (< 10%), the best ratio of gradient refractive index for double ARCs must
174
meet Eq.9. For above reasons, It is found that sample h is optimized double ARCs
ACCEPTED MANUSCRIPT with gradient refractive index in a broader wavelength range compared to other
176
samples. Therefore the reflection spectra simulated by FDTD method in Fig. 5(B)
177
accords very well to the experimental result characterized by UV-visible
178
spectrophotometer in Fig. 5(A), especially in wavelength range from 500 nm to 600
179
nm.
RI PT
175
180
4. CONCLUSIONS
SC
181
In this paper, the single- and double-layer structures of SixOy ARCs with gradient
183
refractive index are numerically and experimentally investigated. The FDTD
184
simulative model is used to predict the optimal reflection spectra corresponding to
185
double ARCs structure of gradient refractive index distribution. The lowest average
186
reflectance (< 10%) in broader wavelength range is achieved by optimized gas ratio of
187
SiH4 to N2O (outer layer 1:3, inner layer 1:1) in the experiment. The optical
188
characterization of thin films with gradient refractive index accords very well to the
189
structural modeling in the simulation.
EP
TE D
M AN U
182
191 192 193
AC C
190
ACKNOWLEDGMENTS The authors gratefully acknowledge the financial support of the project funded
by National Science Foundation of China (Grant no. 61404080 and 51202139).
194 195
Reference
196
[1] A.H. Mahan, Sol. Energy Mater. Sol. Cells, 78 (2003) 299
ACCEPTED MANUSCRIPT [2] G. Fortunato, Thin Solid Films, 296 (1997) 82
198
[3] S.D. Brotherton, Semicond Sci Technol, 10 (1995) 721
199
[4] S. Duttagupta, F. J. Ma, B. Hoex, A. G. Aberle, Sol. Energy Mater. Sol. Cells, 120 (2014) 204
200
[5] N. Balaji, K. Song, J. Choi, C. Park, J. Yi, Curr. Appl. Phys., 13 (2013) 1826
201
[6] H. P. Zhou, D. Y. Wei, L. X. Xu, Y. N. Guo, S. Q. Xiao, S. Y. Huang, S. Xu, Appl. Surf. Sci.,
202
264 (2013) 21
203
[7] Q. Qiao, H. Y. Lu, J. Ge, X. Xi, R. L. Chen, J. Yang, J. B. Zhu, Z. R. Shi, J. H. Chu, Appl. Surf.
204
Sci., 305 (2014) 439
205
[8] H. Lee, T. Tachibana, N. Ikeno, H. Hashiguchi, K. Arafune, H. Yoshida, S. I. Satoh, T.
206
Chikyow, A. Ogura, Appl. Phys. Lett, 100 (2012) 143901
207
[9] T. Dullweber, C. Kranz, B. Beier, B. Veith, J. Schmidt, B. F. P. Roos, O. Hohn, T. Dippell, R.
208
Brendel, Sol. Energy Mater. Sol. Cells, 112 (2013) 196
209
[10] L. Remache, E. Fourmond, A. Mahdjoub, J. Dupuis, M. Lemiti, Mat. Sci. Eng. B: Solid, 176
210
(2011) 45
211
[11] Z. Z. Chen, P. Sana, J. Salami, A. Rohatgi, IEEE T. Electron Dev., 40 (1993) 1161
212
[12] B. He, H. Z. Wang, Y. G. Li, Z. Q. Ma, J. Xu, Q. H. Zhang, C. R. Wang, H. Z. Xing, L. Zhao,
213
Y. C. Rui, J. Alloy. Compd., 581 (2013) 28
214
[13] W. B. Qiu, Y. H. Ma, J. Zhao, J. X. Wang, M. K. Li, S. Y. Li, J. Q. Pan, Japn. J. Appl. Phys.,
215
53 (2014) 021501
216
[14] N. Iari, S. Ahangarani, A. Shanaghi, J. Mater. Eng. Perform., 24 (2015) 2645
217
[15] S. Y. Lien, D. S. Wuu, W. C. Yeh, J. C. Liu, Sol. Energy Mater. Sol. Cells, 90 (2006) 2710
218
[16] H. Y. Li, H. Xiong, Y. X. Tang, L. L. Hu, Rare Metal Mat. Eng., 39 (2010) 145
AC C
EP
TE D
M AN U
SC
RI PT
197
ACCEPTED MANUSCRIPT [17] J. J. Liu, W. J. Ho, Y. Y. Lee, C. M. Chang, Thin Solid Films, 570 (2014) 585
220
[18] D. Z. Li, D. Y. Wan, X. L. Zhu, Y. M. Wang, Z. Han, S. S. Han, Y. K. Shan, F. Q. Huang,
221
Sol. Energy Mater. Sol. Cells, 130 (2014) 505
222
[19] L. Zhao, Y. H. Zuo, C. L. Zhou, H. L. Li, H. W. Diao, W. J. Wang, Solar Energy, 85 (2011)
223
530
224
[20] R. Xu, X. D. Wang, L. Song, W. Liu, A. Ji, F. H. Yang, J. M. Li, Opt. Express, 20 (2012) 5061
225
[21] H. J. Hovel, Semiconductors and Semimetals, Volume 11, Solar Cells (Academic Press, USA,
226
1975)
M AN U
SC
RI PT
219
227 228 229
Table Caption:
TE D
230 231 232
Table 1 Experimental parameters of SixOy:H thin film samples deposited by PECVD system
235
Table 2 The refractive index and thickness of SixOy ARC samples for different gas ratio
AC C
234
EP
233
236
Figure Caption:
237
Fig. 1 The basic cross-section structures of (A) single ARC on the substrate (B) double ARCs on
238
the substrate
239 240
Fig. 2 The characterization of XRD patterns for sample a, b and c
ACCEPTED MANUSCRIPT 241 242
Fig. 3 (A) The cross-section structure of sample h characterized by SEM (B) The weight and
243
atomic percentage of element for sample h characterized by EDS
RI PT
244 245
Fig. 4 The correlation between refractive index and gas ratio of SiH4 to N2O for single SixOy thin
246
films
SC
247
Fig. 5 (A) UV-visible reflection spectra and (B) FDTD simulative reflection spectra of sample a, h,
249
i and j
AC C
EP
TE D
M AN U
248
ACCEPTED MANUSCRIPT Table 1
Specific values
Pressure (Torr)
0.4-0.6
Temperature (°C)
200-250
The Ar2 flow of pretreatment (sccm) Frequency (M Hz)
RI PT
Parameters of deposited SixOy:H thin film
30
13.56
60-80
AC C
EP
TE D
M AN U
SC
RF power (W)
ACCEPTED MANUSCRIPT Table 2 SiH4: N2O: H2 (Vol %)
Refractive index
Thickness (nm)
a b c d
10:5:100 (1:0.5:10) 10:10:100 (1:1:10) 10:20:100 (1:2:10) 8:24:80 (1:3:10) Outer layer: 10:10:100 (1:1:10) Inner layer: 10:5:100 (1:0.5:10) Outer layer: 10:20:100 (1:2:10) Inner layer: 10:5:100 (1:0.5:10) Outer layer: 8:24:80 (1:3:10) Inner layer: 10:20:100 (1:2:10) Outer layer: 8:24:80 (1:3:10) Inner layer: 10:10:100 (1:1:10) Outer layer: 8:24:80 (1:3:10) Inner layer: 10:5:100 (1:0.5:10) Outer layer: 10:20:100 (1:2:10) Inner layer: 10:10:100 (1:1:10)
2.421 2.256 2.013 1.674
95.7 84.8 76.4 67.5
2.287
118.6
g h i
AC C
EP
TE D
j
2.224 1.897 2.064
SC
f
2.105
M AN U
e
RI PT
Sample No.
2.093
107.4 98.5
112.5
105.2 115.7
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
>The SixOy thin film is deposited on the substrate as antireflective coating by PECVD >Different reflection spectra with gradient refractive index are obtained by simulation > Reflection spectra with different structure are measured by UV-visible spectrometer > The characterization of reflective spectra verifies the modeling in FDTD simulation